Question

Difficulty: MediumManaging Storage and Database Solutions

An operations team monitors a high-throughput IoT analytics platform that streams write traffic to a Cloud Bigtable instance. During peak usage windows, monitoring alerts show that CPU utilization on the Bigtable cluster exceeds 80%, causing significant write latency spikes. The team needs an immediate operational resolution to lower CPU usage and restore low latency without application code modifications or data schema changes. Which action should the Cloud Engineer take?

  1. Increase the number of nodes in the Cloud Bigtable cluster using the Cloud Console or gcloud bigtable clusters update command.Answer
  2. B
    Migrate the IoT write workloads to a Cloud SQL for MySQL instance configured with read replicas.
  3. C
    Change the storage class of the Bigtable cluster to Coldline to reduce memory and processing overhead.
  4. D
    Execute gsutil perfdiag against the Bigtable cluster endpoint to automatically optimize IOPS throughput.

Answer

Increase the number of nodes in the Cloud Bigtable cluster using the Cloud Console or gcloud bigtable clusters update command.
Cloud Bigtable scales compute and storage independently. When high CPU utilization causes increased latency, increasing the node count instantly redistributes the workload across more nodes, decreasing per-node CPU utilization and lowering latency with zero downtime.

Step-by-Step Solution

1
Identify the performance constraint from telemetry metrics.
CPU utilization exceeding 80% on Cloud Bigtable directly correlates with write latency spikes.
Cloud Bigtable performance scales linearly with the number of nodes assigned to the cluster.
2
Select an operational action that scales throughput capacity dynamically.
Adding nodes increases computing power and bandwidth allocated to the cluster.
Node scaling in Cloud Bigtable takes effect online without downtime or application modifications.
3
Use the appropriate administrative tool to adjust cluster size.
Execute gcloud bigtable clusters update to adjust node count.
Management operations for Cloud Bigtable clusters are performed via gcloud bigtable commands or the Cloud Console.

Key Concept

Cloud Bigtable Performance Management and Operational Scaling
Rate this question